Building Reliable AI: From Knowledge Work to Data Agents

​For the coming February DataScienceSG Meetup, we're exploring a critical theme: reliability in the age of AI. Join us at Google Developer Space as we examine how to build trustworthy AI systems—from auditable knowledge work to data agents.

​Agenda

​6:30 PM – 7:00 PM Registration & Networking

​7:00 PM – 7:30 PM Can Knowledge Work Be AI-Auditable?

​7:30 PM – 8:00 PM Engineering Reliable Data Agents for the Enterprise

​8:00 PM – 8:30 PM Q&A and Closing

​Synopsis

​Session 1: Can Knowledge Work Be AI-Auditable? Speaker: Loy Hui Chien

​Large Language Models produce answers that are fluent and confident regardless of whether they are well-founded. Unlike buggy code, bad analysis doesn't crash; it quietly misleads. This doesn't go away even as models get better. This talk explores an empirical alternative for complex knowledge questions: a structured workflow that separates planning, evidence gathering, and synthesis, helping make clear what outputs can and cannot establish. As LLM use proliferates in knowledge work, the problem is urgent and the solution space underexplored.

​Session 2: Engineering Reliable Data Agents for the Enterprise Speaker: Thu Ya

​The "Agentic Era" promises AI that can browse databases, write its own code, and execute complex workflows. But for Data Science teams, this transition brings a new set of challenges: reliability, safety, and evaluation. This talk cuts through the hype to examine the technical pillars of Agentic Data Science. We discuss how to move from simple RAG (Retrieval-Augmented Generation) to sophisticated agentic workflows, the necessity of rigorous "Evals" in non-deterministic systems, and why the future of ETL might be a self-healing autonomous loop.

​Speaker Bios

​Loy Hui Chien is a public sector professional currently on a temporary career break. A long-time computing enthusiast, he believes generative AI marks the most exciting era of computing in decades. Through DataScienceSG, he seeks to connect with AI changemakers in Singapore and strengthen our collective AI journey.

​Thu Ya is currently a Senior Developer Advocate at Google. At Google, he helps developers and builders succeed in building and deploying AI Applications on Google Cloud. He is also passionate about sharing his knowledge with the communities.